Learn how AI agents securely manage identity and establish trust across dynamic workflows in this technical presentation by IBM Distinguished Engineer Grant Miller. Explore the critical concepts of identity propagation, token exchange, and transitive trust within agentic systems, understanding how these mechanisms prevent identity impersonation and maintain security in multi-agent environments. Discover the role of OAuth 2 and OpenID Connect protocols in enabling secure authentication and authorization for AI agents, and examine how APIs facilitate trusted communication between different agents in complex system architectures. Gain insights into the challenges and solutions for maintaining identity integrity as AI agents interact across various services and platforms, ensuring that trust relationships remain intact throughout distributed workflows.